library(magrittr)
library(tidyverse)
library(Seurat)
library(readxl)
library(cowplot)
library(colorblindr)
library(viridis)
library(progeny)
library(destiny)

coi <- params$cell_type_super
cell_sort <- params$cell_sort
cell_type_major <- params$cell_type_major
louvain_resolution <- params$louvain_resolution
louvain_cluster <- params$louvain_cluster

1 Cluster markers

1.1 Major Myeloid.super markers for cell assign

### load all data ---------------------------------
source("_src/global_vars.R")

# seu_obj <- read_rds(paste0("/work/shah/isabl_data_lake/analyses/16/52/1652/celltypes/", coi, "_processed.rds"))
seu_obj <- read_rds(paste0("/work/shah/uhlitzf/data/SPECTRUM/freeze/v6/outs_pre/", coi, "_seurat_", louvain_resolution, ".rds"))
# seu_obj <- read_rds(paste0("/work/shah/uhlitzf/data/SPECTRUM/freeze/v6/", coi, "_highqc.rds"))

myfeatures <- c("umapharmony_1", "umapharmony_2", "sample", louvain_cluster, "doublet", "nCount_RNA", "nFeature_RNA", "percent.mt", "doublet_score")

plot_data_wrapper <- function(cluster_res) {
  cluster_res <- enquo(cluster_res)
  as_tibble(FetchData(seu_obj, myfeatures)) %>% 
    left_join(meta_tbl, by = "sample") %>% 
    rename(cluster = !!cluster_res) %>% 
    mutate(cluster = as.character(cluster),
           tumor_supersite = ordered(tumor_supersite, levels = names(clrs$tumor_supersite)))
}

plot_data <- plot_data_wrapper(louvain_cluster)

1.2 Subtype currated markers

helper_f <- function(x) ifelse(is.na(x), "", x)

markers_v6_super[[coi]] %>% 
  group_by(subtype) %>% 
  mutate(rank = row_number(gene)) %>% 
  spread(subtype, gene) %>% 
  mutate_all(.funs = helper_f) %>% 
  formattable::formattable()

rank	cDC1	cDC2	Cycling.M	dissociated	M1.CDKN1C	M1.CXCL10	M1.S100A8	M1.S100A9	M2.MARCO	M2.SELENOP	Mast.cell	mDC	Neutrophil	pDC
1	BATF3	CD1C	CDC20	BTG1	CD52	CCL8	APOBEC3A	AREG	CCL20	A2M	CPA3	BIRC3	CSF3R	BCL11A
2	BTLA	CLEC10A	CDK1	DNAJB1	CDKN1C	CXCL10	FCN1	EREG	CXCL3	CCL3L1	CTSG	CCL19	CXCR2	CLIC3
3	CLEC9A	FCER1A	MKI67	DUSP1	LST1	CXCL11	S100A8	S100A8	FN1	CCL4L2	GATA2	CCR7	FCGR3B	GZMB
4	IRF8	HLA-DQA1	PTTG1	EGR1	SMIM25	IFIT1	S100A9	S100A9	HAMP	FCGBP	HPGD	IDO1	IL1R2	JCHAIN
5	XCR1	S100B	TOP2A	FOS	TCF7L2	TNFSF10	VCAN	THBS1	LYVE1	SELENOP	HPGDS	MARCKSL1	NAMPT	PTGDS
6				FOSB					MARCO		KIT		TXNIP
7				HSPA1A					PLIN2		MS4A2
8				HSPA1B					TGM2		TPSAB1
9				HSPA6							TPSB2
10				JUN
11				JUNB
12				KLF2
13				MT1E
14				MT1G
15				MT1H
16				MT1X

1.3 Subtype cluster markers

# marker_tbl <- read_tsv(paste0("/work/shah/isabl_data_lake/analyses/16/52/1652/celltypes/", coi, "_markers.tsv")) %>% 
#   filter(resolution == louvain_resolution)
marker_tbl <- read_tsv(paste0("/work/shah/uhlitzf/data/SPECTRUM/freeze/v6/outs_pre/", coi, "_markers_", louvain_resolution, ".tsv"))
# marker_tbl <- read_tsv(paste0("/work/shah/uhlitzf/data/SPECTRUM/freeze/v6/", coi, "_highqc_markers_02.tsv"))

## Hypergeometric test --------------------------------------

test_set <- marker_tbl %>% 
  group_by(cluster) %>% 
  filter(!str_detect(gene, "^RPS|^RPL")) %>% 
  slice(1:50) %>% 
  mutate(k = length(cluster)) %>% 
  ungroup %>%
  select(cluster, gene, k) %>% 
  mutate(join_helper = 1) %>% 
  group_by(cluster, join_helper, k) %>%
  nest(test_set = gene)

markers_doub_tbl <- markers_v6 %>% 
  enframe("subtype", "gene") %>% 
  filter(!(subtype %in% unique(c(coi, cell_type_major)))) %>% 
  unnest(gene) %>% 
  group_by(gene) %>% 
  filter(length(gene) == 1) %>% 
  mutate(subtype = paste0("doublet.", subtype)) %>% 
  bind_rows(tibble(subtype = "Mito.high", gene = grep("^MT-", rownames(seu_obj), value = T)))

ref_set <- markers_v6_super[[coi]] %>% 
  bind_rows(markers_doub_tbl) %>% 
  group_by(subtype) %>% 
  mutate(m = length(gene),
         n = length(rownames(seu_obj))-m,
         join_helper = 1) %>% 
  group_by(subtype, m, n, join_helper) %>%
  nest(ref_set = gene)

hyper_tbl <- test_set %>% 
  left_join(ref_set, by = "join_helper") %>% 
  group_by(cluster, subtype, m, n, k) %>%
  do(q = length(intersect(unlist(.$ref_set), unlist(.$test_set)))) %>%
  mutate(pval = 1-phyper(q = q, m = m, n = n, k = k)) %>%
  ungroup %>%
  mutate(qval = p.adjust(pval, "BH"),
         sig = qval < 0.01)

# hyper_tbl %>% 
#   group_by(subtype) %>% 
#   filter(any(qval < 0.01)) %>%
#   ggplot(aes(subtype, -log10(qval), fill = sig)) +
#   geom_bar(stat = "identity") +
#   facet_wrap(~cluster) +
#   coord_flip()
  
low_rank <- str_detect(unique(hyper_tbl$subtype), "doublet|dissociated")
subtype_lvl <- c(sort(unique(hyper_tbl$subtype)[!low_rank]), sort(unique(hyper_tbl$subtype)[low_rank]))
  
cluster_label_tbl <- hyper_tbl %>% 
  mutate(subtype = ordered(subtype, levels = subtype_lvl)) %>% 
  arrange(qval, subtype) %>%
  group_by(cluster) %>% 
  slice(1) %>% 
  mutate(subtype = ifelse(sig, as.character(subtype), paste0("unknown_", cluster))) %>% 
  select(cluster, cluster_label = subtype) %>% 
  ungroup %>% 
  mutate(cluster_label = make.unique(cluster_label, sep = "_"))

seu_obj$cluster_label <- unname(deframe(cluster_label_tbl)[as.character(unlist(seu_obj[[paste0("RNA_snn_res.", louvain_resolution)]]))])
plot_data$cluster_label <- seu_obj$cluster_label

cluster_n_tbl <- seu_obj$cluster_label %>% 
  table() %>% 
  enframe("cluster_label", "cluster_n") %>% 
  mutate(cluster_nrel = cluster_n/sum(cluster_n))

marker_sheet <- marker_tbl %>% 
  left_join(cluster_label_tbl, by = "cluster") %>% 
  group_by(cluster_label) %>% 
  filter(!str_detect(gene, "^RPS|^RPL")) %>% 
  slice(1:50) %>% 
  mutate(rank = row_number(-avg_logFC)) %>% 
  select(cluster_label, gene, rank) %>% 
  ungroup %>% 
  mutate(cluster_label = ordered(cluster_label, levels = unique(c(names(clrs$cluster_label[[coi]]), sort(cluster_label))))) %>% 
  spread(cluster_label, gene) %>% 
  mutate_all(.funs = helper_f)

marker_tbl_annotated <- marker_tbl %>% 
  left_join(cluster_label_tbl, by = "cluster") %>% 
  left_join(cluster_n_tbl, by = "cluster_label") %>% 
  select(-cluster, -resolution) %>% 
  mutate(cluster_label = ordered(cluster_label, levels = unique(c(names(clrs$cluster_label[[coi]]), sort(cluster_label))))) %>% 
  arrange(cluster_label, -avg_logFC, p_val_adj)

write_tsv(marker_sheet, paste0("/work/shah/uhlitzf/data/SPECTRUM/freeze/v6/", coi, "_marker_sheet.tsv"))

write_tsv(marker_tbl_annotated, paste0("/work/shah/uhlitzf/data/SPECTRUM/freeze/v6/", coi, "_marker_table_annotated.tsv"))

formattable::formattable(marker_sheet)

rank	M1.S100A8	M1.CDKN1C	M1.CXCL10	M2.SELENOP	M2.MARCO	cDC1	cDC2	mDC	pDC	Mast.cell	Cycling.M	dissociated	doublet.Fibroblast	doublet.Fibroblast_1	doublet.Ovarian.cancer.cell	doublet.Plasma.cell	doublet.T.cell
1	S100A9	CDKN1C	CXCL10	SELENOP	MARCO	CPVL	FCER1A	BIRC3	GZMB	TPSAB1	STMN1	MT1G	COL1A1	MMP9	SLPI	IGKC	IL32
2	S100A8	SMIM25	CXCL11	SLC40A1	FN1	CLEC9A	S100B	CCL19	PTGDS	TPSB2	H2AFZ	MT1H	COL1A2	CTSK	WFDC2	MTRNR2L12	CCL5
3	VCAN	LST1	CCL8	A2M	HAMP	LGALS2	CLEC10A	CCR7	JCHAIN	CPA3	CENPF	PLIN2	COL3A1	ACP5	KRT19	FTL	CD2
4	FCN1	CD52	IFIT2	DAB2	LYVE1	CPNE3	CD1C	MARCKSL1	CLIC3	HPGD	MKI67	MT1X	CALD1	SLC9B2	LCN2	CXCL8	GZMA
5	S100A12	LILRB2	TNFSF10	CCL3	CXCL3	IDO1	PPA1	TXN	BCL11A	MS4A2	HIST1H4C	MT2A	SPARC	ANPEP	KRT18	NUPR1	CD3D
6	THBS1	FCN1	IFIT1	CCL4L2	TIMP1	CST3	HLA-DPB1	IDO1	ITM2C	CTSG	TOP2A	CXCL8	DCN	SPARC	CLU	IGLC2	NKG7
7	LYZ	CFP	ISG15	APOE	CXCL2	C1orf54	HLA-DQA1	LAMP3	IRF7	CD69	TUBB	MT1E	SELENOP	ANKH	KRT7	APOC1	TRAC
8	EREG	IFITM2	RSAD2	CCL3L1	CCL20	SNX3	CD1E	IL7R	LILRA4	CLU	TUBA1B	HSPA1A	TAGLN	LMNA	KRT8	MTRNR2L8	GNLY
9	AC020656.1	LILRA5	IFIT3	FCGBP	CFD	DNASE1L3	HLA-DQB1	PPA1	TCF4	HPGDS	TYMS	HSPA6	C1QB	CD109	CD24	MARCO	LTB
10	AREG	RIPOR2	GBP1	SLCO2B1	PLTP	WDFY4	HLA-DPA1	CCL22	PLAC8	RHEX	HMGN2	CCL20	C1QA	SIGLEC15	S100A1	CSF3R	TRBC2
11	RETN	APOBEC3A	CXCL9	RNASE1	CTSL	IRF8	AREG	CCL17	TCL1A	KIT	HMGB2	HK2	C1QC	TNFRSF11A	S100A13	APOE	IL7R
12	CD55	LYPD2	APOBEC3A	C1QB	CYP27A1	S100B	JAML	EBI3	IRF8	SLC18A2	HMGB1	ADM	NNMT	ANXA2	IGFBP2	PLCG2	CD3G
13	MCEMP1	TCF7L2	MX1	C1QC	EMP3	RGCC	PKIB	FSCN1	CCDC50	VWA5A	PCLAF	G0S2	FOLR2	VIM	C19orf33	P4HA1	TRBC1
14	CYP1B1	CD55	ISG20	HTRA1	S100A8	CLNK	HLA-DRB1	NUB1	IL3RA	GATA2	NUSAP1	NUPR1	ACTA2	ATP6V0D2	MDK	GPNMB	CD3E
15	CD300E	SPN	CCL2	LGMN	VSIG4	XCR1	HLA-DRA	CSF2RA	TSPAN13	AREG	PTTG1	DNAJB1	F13A1	CD99	EPCAM	SLC25A37	GZMK
16	CD36	COTL1	GBP4	F13A1	CD163	HLA-DPB1	HLA-DQA2	CST7	UGCG	IL1RL1	CKS1B	ERO1A	APOE	MATK	CLDN3	CTSD	SYNE2
17	SLC25A37	LYST	HERC5	C1QA	GLUL	CCND1	HLA-DQB2	DUSP5	PLD4	LTC4S	ASPM	NDRG1	TMEM176B	SUCNR1	MMP7	SORL1	CXCL13
18	FPR1	STXBP2	GBP5	GPR34	S100A9	RAB7B	LGALS2	BASP1	PPP1R14B	HDC	PCNA	SLC2A3	SLC40A1	MMP14	MAL2	MALAT1	ETS1
19	MNDA	KLF2	IFIH1	CCL4	CTSB	NAAA	HLA-DRB5	DAPP1	C12orf75	LMO4	TPX2	CSTB	IGFBP2	HCFC1R1	MUC1	FTH1	CD7
20	SERPINA1	POU2F2	NT5C3A	C3	EREG	HLA-DPA1	CST3	CD83	MZB1	RGS13	SMC4	HSPA1B	COL6A2	MYO1B	ELF3	IGHG1	KLRB1
21	NAMPT	CYTIP	IL4I1	FOLR2	PDK4	BATF3	LTB	GPR157	TPM2	AL157895.1	TK1	BNIP3	GPR34	DPP4	LAPTM4B	AC020656.1	SPOCK2
22	S100A4	WARS	IFI6	STAB1	CD14	HLA-DQB1	PLAC8	HLA-DQA1	DERL3	IER2	CDK1	HILPDA	SLCO2B1	ITGA2	S100A16	APLP2	CLEC2D
23	RGS2	SERPINA1	LY6E	APOC1	SOD2	HLA-DQA1	CSF2RA	PTPN1	LTB	ADCYAP1	DUT	SLC2A1	A2M	COL6A1	CRABP2	IGLC3	CST7
24	S100A6	FGR	MX2	MRC1	LTA4H	TXN	PAK1	DUSP4	SPIB	SOCS1	TUBA1C	CXCL3	LUM	COL6A2	BCAM		TNFRSF18
25	CFP	PECAM1	IL1RN	MAF	SLC2A5	SLAMF7	CKLF	LGALS2	CLEC4C	ANXA1	BIRC5	GAPDH	TPM2	NFATC1	CP		ITM2A
26	CD52	PLAC8	SAMD9L	MS4A7	CD55	PPA1	CD1A	HLA-DPB1	SMPD3	PPP1R15A	NUCKS1	ANGPTL4	C3	FNBP1L	SNCG		IKZF3
27	MXD1	S100A4	PARP14	TREM2	FBP1	TACSTD2	CST7	RFTN1	SERPINF1	BACE2	KPNA2	TIMP1	CTSC	AK5	S100A14		CD69
28	APOBEC3A	IFI30	OAS1	MSR1	KLF2	SHTN1	NDRG2	HLA-DRA	MAP1A	LIF	SMC2	MT1M	RARRES2	MYO1D	RBP1		LCK
29	CCR2	HES4	DDX58	IFI27	VMO1	VMO1	IFITM2	SLCO5A1	IRF4	GRAP2	UBE2C	P4HA1	COL6A1	PPIC	TM4SF1		DUSP4
30	IRS2	CFD	EPSTI1	PLD3	C1QB	ASB2	IGSF6	POGLUT1	SCT	RAC2	RAN	LDHA	DAB2	EMP1	S100A2		TNFRSF4
31	CSF3R	ARL4A	IFITM3	NRP1	CXCL8	KCNK6	CIITA	LAD1	CYB561A3	MAOB	CDKN3	FAM162A	HLA-DRB1	COL27A1	PERP		GZMH
32	RIPOR2	TIMP1	STAT1	AKR1B1	MT2A	RAB32	GAPT	TRAF1	SOX4	CSF1	RRM2	C15orf48	CD74	CAVIN3	DSP		ACAP1
33	IRAK3	LILRA1	USP18	PDK4	C2	CADM1	EEF1B2	AOC1	SELL	RHOH	PARP1	EREG	HLA-DRA	RAI14	MAL		CD27
34	STXBP2	CD300E	OASL	IGF1	TIMD4	CST7	CD1D	BCL2L14	APP	KRT1	DEK	BNIP3L	IGF1	AC092484.1	SMIM22		TIGIT
35	SOCS3	NAAA	WARS	CTSD	SLC39A8	BASP1	PLD4	CLIC2	SPCS1	LAPTM4A	RANBP1	MARCO	MS4A7	PPFIBP1	MUC16		CD8A
36	IL1B	TUBA1A	IFI35	GYPC	C1QA	CD74	GPAT3	SPIB	CCDC186	SLC45A3	NASP	VCAN	ITM2B	PHETA1	CLDN4		CTSW
37	CLEC4E	EMP3	IFI44L	CPM	FABP5	HLA-DRB1	TACSTD2	AL031316.1	LAMP5	LAT	CLSPN	SPP1	MS4A6A	DNM1	MSLN		KLRD1
38	CSTA	ADGRE5	RNF213	HLA-DRB5	FTH1	LSP1	ALDH2	NET1	LINC00996	UBB	HMMR	GPNMB	HLA-DPB1	DIXDC1	CDKN2A		CTLA4
39	FGR	CD48	XAF1	CD81	PSAP	GSTP1	YWHAH	CD274	SEL1L3	LAX1	TUBB4B	MXI1	HLA-DPA1	F2R	ASS1		EVL
40	THBD	TNFRSF1B	SAMD9	SLC1A3	FTL	PTPRE	MS4A6A	PTGIR	IGKC	STX3	CCNB1	CLEC2B	COL6A3	NRIP3	FOLR1		CD247
41	MEGF9	CX3CR1	LAP3	FCGR2A	TFRC	FGL2	EEF1A1	MREG	MAPKAPK2	RAB27B	CEP55	CXCL2	CCDC80	RUFY4	PDZK1IP1		CCND2
42	CD93	LILRB1	VAMP5	EGR2	COLEC12	FNBP1	SPINT2	SLC7A11	SEC61B	SMYD3	H2AFV	VEGFA	POSTN	AKAP6	SCGB2A1		RORA
43	NCF2	MYO1G	TNFSF13B	ITM2B	CRIP1	TAP1	CD74	TVP23A	CXCR3	CNRIP1	SPP1	RGCC	MYL9	C1orf21	KRT17		CCL4
44	SH3BGRL3	HCK	OAS3	CD74	C5AR1	HLA-DRA	COTL1	LIMCH1	GPR183	MLPH	ANP32B	FTH1	CTHRC1	ZNF462	CRYAB		IL2RB
45	SELL	CTSS	TAP1	PLTP	TNFRSF1B	ACTG1	LST1	TIFAB	OFD1	ITM2A	HTRA1	BTG1	BGN	STEAP1	NNMT		CD52
46	HBEGF	NAP1L1	DYNLT1	EGR1	TGM2	EEF1B2	FILIP1L	IRF4	IGHM	AL662860.1	TMPO	HIF1A-AS2	NBL1	LAT	MYL9		GZMM
47	LTA4H	CLEC12A	C15orf48	GPNMB	MT1E	NAP1L1	HLA-DMA	SLC41A2	SLC15A4	CAVIN2	ATAD2	MT1F	THY1	HTRA3	SLC34A2		APOBEC3G
48	LILRA5	VSIR	OAS2	PLAU	NCF1	DUSP4	BASP1	AUTS2	PLP2	CALB2	ARL6IP1	MIF	MMP11	F5	CRIP2		CD96
49	DUSP1	FLNA	APOL6	OLFML3	VAMP5	SUB1	LIMD2	ANKRD33B	SELENOS	ITM2C	DTYMK	VIM	CXCL12	CTTN	ARHGAP29		RASGRP1
50	JAML	LRRC25	CMPK2	AP1B1	APLP2	ICAM3	LSP1	TNFRSF9	TXN	TIMP3	LDHB	SLC2A5	AEBP1	SLC25A4	CTTN		STK17A

2 Clusters

2.1 sizes

enframe(sort(table(seu_obj$cluster_label))) %>% 
  mutate(name = ordered(name, levels = rev(name))) %>% 
  ggplot() +
  geom_bar(aes(name, value), stat = "identity") +
  theme(axis.text.x = element_text(angle = 90, hjust = 1, vjust = 0.5)) +
  labs(y = c("#cells"), x = "cluster")

2.2 UMAP

alpha_lvl <- ifelse(nrow(plot_data) < 20000, 0.2, 0.1)
pt_size <- ifelse(nrow(plot_data) < 20000, 0.2, 0.05)

common_layers_disc <- list(  
  geom_point(size = pt_size, alpha = alpha_lvl),
  NoAxes(),
  guides(color = guide_legend(override.aes = list(size = 2, alpha = 1))),
  labs(color = "")
)

common_layers_cont <- list(  
  geom_point(size = pt_size, alpha = alpha_lvl),
  NoAxes(),
  scale_color_gradientn(colors = viridis(9)),
  guides(color = guide_colorbar())
)

ggplot(plot_data, aes(umapharmony_1, umapharmony_2, color = cluster_label)) + 
  common_layers_disc +
  #facet_wrap(~therapy) +
  ggtitle("Sub cluster")

3 Filtering out doublet and dissociated clusters

3.1 UMAP

my_subtypes <- names(clrs$cluster_label[[coi]])
my_subtypes <- c(my_subtypes, unlist(lapply(paste0("_", 1:3), function(x) paste0(my_subtypes, x)))) %>% .[!str_detect(., "doublet|dissociated")]
my_subtypes <- my_subtypes[my_subtypes %in% unique(seu_obj$cluster_label)]
my_subtypes <- my_subtypes[my_subtypes %in% names(clrs$cluster_label[[coi]])]

cells_to_keep <- colnames(seu_obj)[seu_obj$cluster_label %in% my_subtypes]
# seu_obj_sub <- subset(seu_obj, cells = cells_to_keep)
# seu_obj_sub <- RunUMAP(seu_obj_sub, dims = 1:50, reduction = "harmony", reduction.name = "umapharmony", reduction.key = "umapharmony_")
# seu_obj_sub$cluster_label <- seu_obj$cluster_label[colnames(seu_obj) %in% colnames(seu_obj_sub)]
# write_rds(seu_obj_sub, paste0("/work/shah/uhlitzf/data/SPECTRUM/freeze/v6/", coi, "_processed_filtered.rds"))
seu_obj_sub <- read_rds(paste0("/work/shah/uhlitzf/data/SPECTRUM/freeze/v6/", coi, "_processed_filtered.rds"))

plot_data_sub <- as_tibble(FetchData(seu_obj_sub, c(myfeatures, "cluster_label"))) %>% 
  left_join(meta_tbl, by = "sample") %>% 
  mutate(patient_id = str_remove_all(patient_id, "SPECTRUM-OV-"),
         tumor_supersite = ordered(tumor_supersite, levels = names(clrs$tumor_supersite))) %>% 
  mutate(cell_id = colnames(seu_obj_sub),
         cluster_label = ordered(cluster_label, levels = my_subtypes),
         )
  
if (cell_sort == "CD45+") {
plot_data_sub <- filter(plot_data_sub, sort_parameters != "singlet, live, CD45-", !is.na(tumor_supersite))
}

if (cell_sort == "CD45-") {
plot_data_sub <- filter(plot_data_sub, sort_parameters != "singlet, live, CD45+", !is.na(tumor_supersite))
}

ggplot(plot_data_sub, aes(umapharmony_1, umapharmony_2, color = cluster_label)) + 
  common_layers_disc +
  ggtitle("Sub cluster") +
  #facet_wrap(~cluster_label) +
  scale_color_manual(values = clrs$cluster_label[[coi]])

ggplot(plot_data_sub, aes(umapharmony_1, umapharmony_2, color = patient_id_short)) + 
  common_layers_disc +
  ggtitle("Patient") +
  # facet_wrap(~therapy) +
  scale_color_manual(values = clrs$patient_id_short)

ggplot(plot_data_sub, aes(umapharmony_1, umapharmony_2, color = tumor_supersite)) + 
  # geom_point(aes(umapharmony_1, umapharmony_2), 
  #            color = "grey90", size = 0.01, 
  #            data = select(plot_data_sub, -tumor_supersite)) +
  common_layers_disc +
  ggtitle("Site") +
  # facet_wrap(~therapy) +
  scale_color_manual(values = clrs$tumor_supersite)

write_tsv(select(plot_data_sub, cell_id, everything(), -umapharmony_1, -umapharmony_2, -contains("RNA_")), paste0("/work/shah/uhlitzf/data/SPECTRUM/freeze/v6/", coi, "_embedding.tsv"))

3.2 add module scores and pathway scores

signature_modules_cd8 <- read_excel("_data/small/signatures/SPECTRUM v5 sub cluster markers.xlsx", sheet = 2, skip = 1, range = "M2:P100") %>% 
  gather(module, gene) %>% 
  na.omit() %>% 
  group_by(module) %>% 
  do(gene = c(.$gene)) %>% 
  {setNames(.$gene, .$module)} %>% 
  setNames(paste0(names(.), ".module"))

signature_modules_cin <- yaml::read_yaml("_data/small/signatures/cgas_sting.yaml") %>% 
  .[lapply(., length) > 1] %>% 
  setNames(paste0(names(.), ".module"))

# ## compute expression module scores
# for (i in 1:length(signature_modules_cd8)) {
#   seu_obj_sub <- AddModuleScore(seu_obj_sub, features = signature_modules_cd8[i], name = names(signature_modules_cd8)[i])
#   seu_obj_sub[[names(signature_modules_cd8)[i]]] <- seu_obj_sub[[paste0(names(signature_modules_cd8)[i], "1")]]
#   seu_obj_sub[[paste0(names(signature_modules_cd8)[i], "1")]] <- NULL
#   print(paste(names(signature_modules_cd8)[i], "DONE"))
# }
# 
# for (i in 1:length(signature_modules_cin)) {
#   seu_obj_sub <- AddModuleScore(seu_obj_sub, features = signature_modules_cin[i], name = names(signature_modules_cin)[i])
#   seu_obj_sub[[names(signature_modules_cin)[i]]] <- seu_obj_sub[[paste0(names(signature_modules_cin)[i], "1")]]
#   seu_obj_sub[[paste0(names(signature_modules_cin)[i], "1")]] <- NULL
#   print(paste(names(signature_modules_cin)[i], "DONE"))
# }
# 
# 
# ## compute progeny scores
# progeny_list <- seu_obj_sub@assays$RNA@data[VariableFeatures(seu_obj_sub),] %>%
#   as.matrix %>%
#   progeny %>%
#   as.data.frame %>%
#   as.list
# 
# names(progeny_list) <- make.names(paste0(names(progeny_list), ".pathway"))
# 
# for (i in 1:length(progeny_list)) {
#   seu_obj_sub <- AddMetaData(seu_obj_sub,
#                                  metadata = progeny_list[[i]],
#                                  col.name = names(progeny_list)[i])
# }
# 
# write_rds(seu_obj_sub, paste0("/work/shah/uhlitzf/data/SPECTRUM/freeze/v6/", coi, "_processed_filtered_annotated.rds"))
seu_obj_sub <- read_rds(paste0("/work/shah/uhlitzf/data/SPECTRUM/freeze/v6/", coi, "_processed_filtered_annotated.rds"))

3.3 marker heatmap

marker_top_tbl <- marker_sheet[,-1] %>% 
  slice(1:10) %>% 
  as.list %>% 
  .[!str_detect(names(.), "doublet")] %>% 
  enframe("cluster_label_x", "gene") %>% 
  unnest(gene)

plot_data_markers <- as_tibble(FetchData(seu_obj_sub, c("cluster_label", myfeatures, unique(marker_top_tbl$gene)))) %>% 
  gather(gene, value, -c(1:(length(myfeatures)+1))) %>% 
  left_join(meta_tbl, by = "sample") %>% 
  mutate(patient_id = str_remove_all(patient_id, "SPECTRUM-OV-"),
         tumor_supersite = ordered(tumor_supersite, levels = names(clrs$tumor_supersite))) %>% 
  mutate(cluster_label = ordered(cluster_label, levels = my_subtypes)) %>% 
  group_by(cluster_label, gene) %>% 
  summarise(value = mean(value, na.rm = T)) %>% 
  group_by(gene) %>% 
  mutate(value = scales::rescale(value)) %>% 
  left_join(marker_top_tbl, by = "gene") %>% 
  mutate(cluster_label_x = ordered(cluster_label_x, levels = rev(names(clrs$cluster_label[[coi]]))))

ggplot(plot_data_markers) +
  geom_tile(aes(gene, cluster_label, fill = value)) +
  theme(axis.text.x = element_text(angle = 90, hjust = 1, vjust = 0.5)) +
  facet_grid(~cluster_label_x, scales = "free", space = "free") +
  scale_fill_gradientn(colors = viridis(9)) +
  labs(fill = "Scaled\nexpression") +
  theme(aspect.ratio = 1,
        axis.line = element_blank(),
        axis.ticks = element_blank(),
        axis.title = element_blank())

# ggsave(paste0("_fig/002_marker_heatmap_", coi, ".pdf"), width = nrow(marker_top_tbl)/6, height = 5)

3.4 composition

3.4.1 per site

comp_site_tbl <- plot_data_sub %>%
  filter(!is.na(tumor_supersite)) %>% 
  group_by(cluster_label, tumor_supersite) %>%
  tally %>%
  group_by(tumor_supersite) %>%
  mutate(nrel = n/sum(n)*100) %>%
  ungroup

pnrel_site <- ggplot(comp_site_tbl) +
  geom_bar(aes(tumor_supersite, nrel, fill = cluster_label),
           stat = "identity", position = position_stack()) +
  scale_y_continuous(expand = c(0, 0)) +
  theme(axis.text.x = element_text(angle = 90, hjust = 1, vjust = 0.5)) +
  labs(fill = "Cluster", y = "Fraction [%]", x = "") +
  scale_fill_manual(values = clrs$cluster_label[[coi]])

pnabs_site <- ggplot(comp_site_tbl) +
  geom_bar(aes(tumor_supersite, n, fill = cluster_label),
           stat = "identity", position = position_stack()) +
  scale_y_continuous(expand = c(0, 0)) +
  theme(axis.text.x = element_text(angle = 90, hjust = 1, vjust = 0.5)) +
  labs(fill = "Cluster", y = "# cells", x = "") +
  scale_fill_manual(values = clrs$cluster_label[[coi]])

plot_grid(pnabs_site, pnrel_site, ncol = 2, align = "h")

# ggsave(paste0("_fig/02_deep_dive_", coi, "_comp_site.pdf"), width = 8, height = 4)

3.4.2 per sample

comp_tbl_sample_sort <- plot_data_sub %>% 
  group_by(tumor_subsite, tumor_supersite, patient_id, consensus_signature, therapy, cluster_label) %>% 
  tally %>% 
  group_by(tumor_subsite, tumor_supersite, patient_id, consensus_signature, therapy) %>% 
  mutate(nrel = n/sum(n)*100,
         nsum = sum(n),
         log10n = log10(n),
         label_supersite = "Site",
         label_mutsig = "Signature",
         label_therapy = "Rx") %>% 
  ungroup %>% 
  arrange(desc(therapy), tumor_supersite) %>% 
  mutate(tumor_subsite_rx = paste0(tumor_subsite, "_", therapy)) %>% 
  mutate(tumor_subsite = ordered(tumor_subsite, levels = unique(tumor_subsite)),
         tumor_subsite_rx = ordered(tumor_subsite_rx, levels = unique(tumor_subsite_rx))) %>% 
  arrange(patient_id) %>% 
  mutate(label_patient_id = ifelse(as.logical(as.numeric(fct_inorder(as.character(patient_id)))%%2), "Patient1", "Patient2"))

sample_id_x_tbl <- plot_data_sub %>% 
  mutate(sort_short_x = cell_sort) %>% 
  distinct(patient_id, sort_short_x, tumor_subsite, therapy, sample) %>% 
  unite(sample_id_x, patient_id, sort_short_x, tumor_subsite, therapy) %>% 
  arrange(sample_id_x)

comp_tbl_sample_sort %>% 
  mutate(sort_short_x = cell_sort) %>% 
  unite(sample_id_x, patient_id, sort_short_x, tumor_subsite_rx) %>% 
  select(sample_id_x, cluster_label, n, nrel, nsum) %>% 
  left_join(sample_id_x_tbl, by = "sample_id_x") %>% 
  write_tsv(paste0("/work/shah/uhlitzf/data/SPECTRUM/freeze/v6/", coi, "_subtype_compositions.tsv"))

ybreaks <- c(500, 1000, 2000, 4000, 6000, 8000, 10000, 15000, 20000)

max_cells_per_sample <- max(comp_tbl_sample_sort$nsum)
ymaxn <- ybreaks[max_cells_per_sample < ybreaks][1]

comp_plot_wrapper <- function(y = "nrel", switch = NULL) {
  if (y == "nrel") ylab <- paste0("Fraction\nof cells [%]")
  if (y == "n") ylab <- paste0("Number\nof cells")
  p <- ggplot(comp_tbl_sample_sort, 
              aes_string("tumor_subsite_rx", y, fill = "cluster_label")) + 
    facet_grid(~patient_id, space = "free", scales = "free", switch = switch) +
    coord_cartesian(clip = "off") + 
    scale_fill_manual(values = clrs$cluster_label[[coi]]) + 
    theme(axis.text.x = element_blank(),
          axis.title.y = element_text(angle = 0, vjust = 0.5, hjust = 0.5, 
                                      margin = margin(0, -0.4, 0, 0, unit = "npc")),
          axis.ticks.x = element_blank(),
          axis.title.x = element_blank(),
          axis.line.x = element_blank(),
          strip.text.y = element_blank(),
          strip.text.x = element_blank(),
          strip.background.y = element_blank(),
          strip.background.x = element_blank(),
          plot.margin = margin(0, 0, 0, 0, "npc")) + 
    labs(x = "",
         y = ylab) +
    guides(fill = FALSE)
  if (y == "nrel") p <- p + 
    geom_bar(stat = "identity") +
    scale_y_continuous(expand = c(0, 0), 
                       breaks = c(0, 50, 100), 
                       labels = c("0", "50", "100"))
  if (y == "n") p <- p + 
    geom_bar(stat = "identity", position = position_stack()) +
    scale_y_continuous(expand = c(0, 0), 
                       breaks = c(0, ymaxn/2, ymaxn),
                       limits = c(0, ymaxn),
                       labels = c("", ymaxn/2, ymaxn)) +
    expand_limits(y = c(0, ymaxn)) +
    theme(panel.grid.major.y = element_line(linetype = 1, color = "grey90", size = 0.5))
  return(p)
} 

common_label_layers <- list(
  geom_tile(color = "white", size = 0.15),
  theme(axis.text.x = element_blank(),
        axis.ticks = element_blank(),
        axis.title.x = element_blank(),
        axis.line.x = element_blank(),
        strip.text = element_blank(),
        strip.background = element_blank(),
        plot.margin = margin(0, 0, 0, 0, "npc")),
  scale_y_discrete(expand = c(0, 0)),
  labs(y = ""),
  guides(fill = FALSE),
  facet_grid(~patient_id, 
             space = "free", scales = "free")
)

comp_label_site <- ggplot(distinct(comp_tbl_sample_sort, tumor_subsite_rx, therapy, tumor_supersite, label_supersite, patient_id), 
       aes(tumor_subsite_rx, label_supersite, 
           fill = tumor_supersite)) + 
  scale_fill_manual(values = clrs$tumor_supersite) +
  common_label_layers

comp_label_rx <- ggplot(distinct(comp_tbl_sample_sort, tumor_subsite_rx, therapy, tumor_supersite, label_therapy, consensus_signature, patient_id), 
       aes(tumor_subsite_rx, label_therapy, 
           fill = therapy)) + 
  scale_fill_manual(values = c(`post-Rx` = "gold3", `pre-Rx` = "steelblue")) +
  common_label_layers

comp_label_mutsig <- ggplot(distinct(comp_tbl_sample_sort, tumor_subsite_rx, therapy, tumor_supersite, label_mutsig, consensus_signature, patient_id), 
       aes(tumor_subsite_rx, label_mutsig, 
           fill = consensus_signature)) + 
  scale_fill_manual(values = clrs$consensus_signature) +
  common_label_layers

patient_label_tbl <- distinct(comp_tbl_sample_sort, patient_id, .keep_all = T)

comp_label_patient_id <- ggplot(patient_label_tbl, aes(tumor_subsite_rx, label_patient_id)) + 
  scale_fill_manual(values = clrs$consensus_signature) +
  geom_text(aes(tumor_subsite_rx, label_patient_id, label = patient_id)) +
  facet_grid(~patient_id, 
             space = "free", scales = "free") +
  coord_cartesian(clip = "off") + 
  theme_void() +
  theme(strip.text = element_blank(),
        strip.background = element_blank(),
        plot.margin = margin(0, 0, 0, 0, "npc"))

hist_plot_wrapper <- function(x = "nrel") {
  if (x == "nrel") {
    xlab <- paste0("Fraction of cells [%]")
    bw <- 5
  }
  if (x == "log10n") {
    xlab <- paste0("Number of cells")
    bw <- 0.2
  }
  p <- ggplot(comp_tbl_sample_sort) +
    ggridges::geom_density_ridges(
      aes_string(x, "cluster_label", fill = "cluster_label"), color = "black",
      stat = "binline", binwidth = bw, scale = 3) +
    facet_grid(label_supersite~., 
               space = "free", scales = "free") +
    scale_fill_manual(values = clrs$cluster_label[[coi]]) + 
    theme(axis.text.y = element_blank(),
          axis.ticks.y = element_blank(),
          axis.title.y = element_blank(),
          axis.line.y = element_blank(),
          strip.text = element_blank(),
          strip.background = element_blank(),
          plot.margin = margin(0, 0, 0, 0, "npc")) +
    scale_x_continuous(expand = c(0, 0)) +
    scale_y_discrete(expand = c(0, 0)) +
    guides(fill = FALSE) +
    labs(x = xlab)
  if (x == "log10n") p <- p + expand_limits(x = c(0, 3)) + 
    scale_x_continuous(expand = c(0, 0), 
                       labels = function(x) parse(text = paste("10^", x)))
  return(p)
}

pcomp1 <- comp_plot_wrapper("n")
pcomp2 <- comp_plot_wrapper("nrel")

pcomp_grid <- plot_grid(comp_label_patient_id, 
                        pcomp1, pcomp2, 
                        comp_label_site, comp_label_rx, comp_label_mutsig,
                        ncol = 1, align = "v", 
                        rel_heights = c(0.15, 0.33, 0.33, 0.06, 0.06, 0.06))

phist1 <- hist_plot_wrapper("log10n")

pcomp_hist_grid <- ggdraw() +
  draw_plot(pcomp_grid, x = 0.01, y = 0, width = 0.85, height = 1) +
  draw_plot(phist1, x = 0.87, y = 0.05, width = 0.12, height = 0.8)

pcomp_hist_grid

# ggsave(paste0("_fig/02_composition_v6_",coi,".pdf"), pcomp_hist_grid, width = 10, height = 2)

3.4.3 site specific cluster enrichment

comp_tbl_z <- comp_tbl_sample_sort %>% 
  filter(therapy == "pre-Rx",
         !(tumor_supersite %in% c("Ascites", "Other"))) %>% 
  group_by(patient_id, cluster_label) %>% 
  arrange(patient_id, cluster_label, nrel) %>% 
  mutate(rank = row_number(nrel),
         z_rank = scales::rescale(rank)) %>% 
  mutate(mean_nrel = mean(nrel, na.rm = T),
         sd_nrel = sd(nrel, na.rm = T),
         z_nrel = (nrel - mean_nrel) / sd_nrel) %>% 
  ungroup()

ggplot(comp_tbl_z) +
  geom_boxplot(aes(tumor_supersite, z_nrel, color = tumor_supersite), 
               outlier.shape = NA) +
  geom_point(aes(tumor_supersite, z_nrel, color = tumor_supersite), position = position_jitter(), size = 0.1) +
  facet_grid(~cluster_label, scales = "free_x", space = "free_x") +
  theme(axis.text.x = element_text(angle = 90, hjust = 1, vjust = 0.5),
        aspect.ratio = 5) +
  scale_color_manual(values = clrs$tumor_supersite)

ggplot(comp_tbl_z) +
  geom_boxplot(aes(tumor_supersite, z_rank, color = tumor_supersite), 
               outlier.shape = NA) +
  geom_point(aes(tumor_supersite, z_rank, color = tumor_supersite), position = position_jitter(), size = 0.1) +
  facet_grid(~cluster_label, scales = "free_x", space = "free_x") +
  theme(axis.text.x = element_text(angle = 90, hjust = 1, vjust = 0.5),
        aspect.ratio = 5) +
  scale_color_manual(values = clrs$tumor_supersite)

3.4.4 macrophage diffusion maps

## subset macrophages
# seu_obj_mp <- seu_obj_sub %>%
#   subset(subset = cluster_label %in% grep("^M1|^M2", unique(seu_obj_sub$cluster_label), value = T)) %>%
#   RunUMAP(dims = 1:50, reduction = "harmony", reduction.name = "umapharmony", reduction.key = "umapharmony_")
# 
# write_rds(seu_obj_mp, "/work/shah/uhlitzf/data/SPECTRUM/freeze/v6/Macrophages_processed.rds")

## subset DCs
seu_obj_dc <- seu_obj_sub %>%
  subset(subset = cluster_label %in% grep("DC|Mast", unique(seu_obj_sub$cluster_label), value = T)) %>%
  RunUMAP(dims = 1:50, reduction = "harmony", reduction.name = "umapharmony", reduction.key = "umapharmony_")

write_rds(seu_obj_dc, "/work/shah/uhlitzf/data/SPECTRUM/freeze/v6/DCs_processed.rds")

seu_obj_dc <- read_rds("/work/shah/uhlitzf/data/SPECTRUM/freeze/v6/DCs_processed.rds")

seu_obj_mp <- read_rds("/work/shah/uhlitzf/data/SPECTRUM/freeze/v6/Macrophages_processed.rds")

# root_cell <- "SPECTRUM-OV-031_S1_CD45P_LEFT_FALLOPIAN_TUBE_AACAAGATCATTGCTT"
# seu_obj_mp_sub <- subset(seu_obj_mp, cells = c(root_cell, colnames(seu_obj_mp)[colnames(seu_obj_mp)!=root_cell][-1]))
# 
# dc_obj <- DiffusionMap(seu_obj_mp_sub@reductions$harmony@cell.embeddings, k = 100)
# dc_mat <- dc_obj@eigenvectors
# colnames(dc_mat) <- paste0("DC_", 1:ncol(dc_mat))
# seu_obj_mp_sub[["DC"]] <- CreateDimReducObject(embeddings = dc_mat, key = "DC_", assay = DefaultAssay(seu_obj_mp_sub))
# 
# write_rds(seu_obj_mp_sub, "/work/shah/uhlitzf/data/SPECTRUM/freeze/v6/Macrophages_processed_filtered.rds")
# 
# dpt_obj <- DPT(dc_obj)
# write_rds(dpt_obj, "/work/shah/uhlitzf/data/SPECTRUM/freeze/v6/Macrophages_sub_dpt.rds")
# seu_obj_mp_sub$DPT1 <- dpt_obj$DPT1
# 
# write_rds(seu_obj_mp_sub, "/work/shah/uhlitzf/data/SPECTRUM/freeze/v6/Macrophages_processed_filtered.rds")
seu_obj_mp_sub <- read_rds("/work/shah/uhlitzf/data/SPECTRUM/freeze/v6/Macrophages_processed_filtered.rds")

# plot_data_sub_sub <- as_tibble(FetchData(seu_obj_sub, c(myfeatures, "cluster_label"))) %>% 
#   left_join(meta_tbl, by = "sample") %>% 
#   mutate(patient_id = str_remove_all(patient_id, "SPECTRUM-OV-"),
#          tumor_supersite = ordered(tumor_supersite, levels = names(clrs$tumor_supersite))) %>% 
#   mutate(cell_id = colnames(seu_obj_sub),
#          cluster_label = ordered(cluster_label, levels = my_subtypes),
#          )
#   
# if (cell_sort == "CD45+") {
# plot_data_sub_sub <- filter(plot_data_sub_sub, sort_parameters != "singlet, live, CD45-", !is.na(tumor_supersite))
# }
# 
# if (cell_sort == "CD45-") {
# plot_data_sub_sub <- filter(plot_data_sub_sub, sort_parameters != "singlet, live, CD45+", !is.na(tumor_supersite))
# }
# 
# ggplot(plot_data_sub_sub, aes(umapharmony_1, umapharmony_2, color = cluster_label)) + 
#   common_layers_disc +
#   ggtitle("Sub cluster") +
#   #facet_wrap(~cluster_label) +
#   scale_color_manual(values = clrs$cluster_label[[coi]])
# 
# ggplot(plot_data_sub_sub, aes(umapharmony_1, umapharmony_2, color = patient_id_short)) + 
#   common_layers_disc +
#   ggtitle("Patient") +
#   # facet_wrap(~therapy) +
#   scale_color_manual(values = clrs$patient_id_short)
# 
# ggplot(plot_data_sub_sub, aes(umapharmony_1, umapharmony_2, color = tumor_supersite)) + 
#   # geom_point(aes(umapharmony_1, umapharmony_2), 
#   #            color = "grey90", size = 0.01, 
#   #            data = select(plot_data_sub_sub, -tumor_supersite)) +
#   common_layers_disc +
#   ggtitle("Site") +
#   # facet_wrap(~therapy) +
#   scale_color_manual(values = clrs$tumor_supersite)
# 
# write_tsv(select(plot_data_sub_sub, cell_id, everything(), -umapharmony_1, -umapharmony_2, -contains("RNA_")), paste0("/work/shah/uhlitzf/data/SPECTRUM/freeze/v6/", coi, "_embedding_sub.tsv"))

3.4.5 UMAP & DC

plot_data_mp_sub <- as_tibble(FetchData(seu_obj_mp_sub, c(myfeatures, "cluster_label", "DC_1", "DC_2"))) %>% 
  left_join(meta_tbl, by = "sample") %>% 
  mutate(patient_id = str_remove_all(patient_id, "SPECTRUM-OV-"),
         tumor_supersite = ordered(tumor_supersite, levels = names(clrs$tumor_supersite))) %>% 
  mutate(cell_id = colnames(seu_obj_mp_sub),
         cluster_label = ordered(cluster_label, levels = names(clrs$cluster_label[[coi]])))
  
if (cell_sort == "CD45+") {
plot_data_mp_sub <- filter(plot_data_mp_sub, sort_parameters != "singlet, live, CD45-", !is.na(tumor_supersite))
}

if (cell_sort == "CD45-") {
plot_data_mp_sub <- filter(plot_data_mp_sub, sort_parameters != "singlet, live, CD45+", !is.na(tumor_supersite))
}

ggplot(plot_data_mp_sub, aes(umapharmony_1, umapharmony_2, color = cluster_label)) + 
  common_layers_disc +
  ggtitle("Sub cluster") +
  #facet_wrap(~cluster_label) +
  scale_color_manual(values = clrs$cluster_label[[coi]])

ggplot(plot_data_mp_sub, aes(umapharmony_1, umapharmony_2, color = patient_id_short)) + 
  common_layers_disc +
  ggtitle("Patient") +
  # facet_wrap(~therapy) +
  scale_color_manual(values = clrs$patient_id_short)

ggplot(plot_data_mp_sub, aes(umapharmony_1, umapharmony_2, color = tumor_supersite)) + 
  common_layers_disc +
  ggtitle("Site") +
  # facet_wrap(~therapy) +
  scale_color_manual(values = clrs$tumor_supersite)

ggplot(plot_data_mp_sub, aes(DC_1, DC_2, color = cluster_label)) + 
  common_layers_disc +
  ggtitle("Sub cluster") +
  #facet_wrap(~cluster_label) +
  scale_color_manual(values = clrs$cluster_label[[coi]])

ggplot(plot_data_mp_sub, aes(DC_1, DC_2, color = patient_id_short)) + 
  common_layers_disc +
  ggtitle("Patient") +
  # facet_wrap(~therapy) +
  scale_color_manual(values = clrs$patient_id_short)

ggplot(plot_data_mp_sub, aes(DC_1, DC_2, color = tumor_supersite)) + 
  common_layers_disc +
  ggtitle("Site") +
  # facet_wrap(~therapy) +
  scale_color_manual(values = clrs$tumor_supersite)

4 session info

devtools::session_info()

## ─ Session info ───────────────────────────────────────────────────────────────
##  setting  value                       
##  version  R version 3.6.2 (2019-12-12)
##  os       Debian GNU/Linux 10 (buster)
##  system   x86_64, linux-gnu           
##  ui       X11                         
##  language (EN)                        
##  collate  en_US.UTF-8                 
##  ctype    en_US.UTF-8                 
##  tz       Etc/UTC                     
##  date     2021-03-01                  
## 
## ─ Packages ───────────────────────────────────────────────────────────────────
##  package              * version    date       lib
##  abind                  1.4-5      2016-07-21 [2]
##  ape                    5.3        2019-03-17 [2]
##  assertthat             0.2.1      2019-03-21 [2]
##  backports              1.1.10     2020-09-15 [1]
##  bibtex                 0.4.2.2    2020-01-02 [2]
##  Biobase                2.46.0     2019-10-29 [2]
##  BiocGenerics           0.32.0     2019-10-29 [2]
##  BiocParallel           1.20.1     2019-12-21 [2]
##  bitops                 1.0-6      2013-08-17 [2]
##  boot                   1.3-24     2019-12-20 [3]
##  broom                  0.7.2      2020-10-20 [1]
##  callr                  3.4.2      2020-02-12 [1]
##  car                    3.0-8      2020-05-21 [1]
##  carData                3.0-4      2020-05-22 [1]
##  caTools                1.17.1.4   2020-01-13 [2]
##  cellranger             1.1.0      2016-07-27 [2]
##  class                  7.3-15     2019-01-01 [3]
##  cli                    2.0.2      2020-02-28 [1]
##  cluster                2.1.0      2019-06-19 [3]
##  codetools              0.2-16     2018-12-24 [3]
##  colorblindr          * 0.1.0      2020-01-13 [2]
##  colorspace           * 1.4-2      2019-12-29 [2]
##  cowplot              * 1.0.0      2019-07-11 [2]
##  crayon                 1.3.4      2017-09-16 [1]
##  curl                   4.3        2019-12-02 [2]
##  data.table             1.12.8     2019-12-09 [2]
##  DBI                    1.1.0      2019-12-15 [2]
##  dbplyr                 2.0.0      2020-11-03 [1]
##  DelayedArray           0.12.2     2020-01-06 [2]
##  DEoptimR               1.0-8      2016-11-19 [1]
##  desc                   1.2.0      2018-05-01 [2]
##  destiny              * 3.0.1      2020-01-16 [1]
##  devtools               2.2.1      2019-09-24 [2]
##  digest                 0.6.25     2020-02-23 [1]
##  dplyr                * 1.0.2      2020-08-18 [1]
##  e1071                  1.7-3      2019-11-26 [1]
##  ellipsis               0.3.1      2020-05-15 [1]
##  evaluate               0.14       2019-05-28 [2]
##  fansi                  0.4.1      2020-01-08 [2]
##  farver                 2.0.3      2020-01-16 [1]
##  fitdistrplus           1.0-14     2019-01-23 [2]
##  forcats              * 0.5.0      2020-03-01 [1]
##  foreign                0.8-74     2019-12-26 [3]
##  formattable            0.2.0.1    2016-08-05 [1]
##  fs                     1.5.0      2020-07-31 [1]
##  future                 1.15.1     2019-11-25 [2]
##  future.apply           1.4.0      2020-01-07 [2]
##  gbRd                   0.4-11     2012-10-01 [2]
##  gdata                  2.18.0     2017-06-06 [2]
##  generics               0.0.2      2018-11-29 [2]
##  GenomeInfoDb           1.22.0     2019-10-29 [2]
##  GenomeInfoDbData       1.2.2      2020-01-14 [2]
##  GenomicRanges          1.38.0     2019-10-29 [2]
##  ggplot.multistats      1.0.0      2019-10-28 [1]
##  ggplot2              * 3.3.2      2020-06-19 [1]
##  ggrepel                0.8.1      2019-05-07 [2]
##  ggridges               0.5.2      2020-01-12 [2]
##  ggthemes               4.2.0      2019-05-13 [1]
##  globals                0.12.5     2019-12-07 [2]
##  glue                   1.3.2      2020-03-12 [1]
##  gplots                 3.0.1.2    2020-01-11 [2]
##  gridExtra              2.3        2017-09-09 [2]
##  gtable                 0.3.0      2019-03-25 [2]
##  gtools                 3.8.1      2018-06-26 [2]
##  haven                  2.3.1      2020-06-01 [1]
##  hexbin                 1.28.0     2019-11-11 [2]
##  hms                    0.5.3      2020-01-08 [1]
##  htmltools              0.5.1.1    2021-01-22 [1]
##  htmlwidgets            1.5.1      2019-10-08 [2]
##  httr                   1.4.2      2020-07-20 [1]
##  ica                    1.0-2      2018-05-24 [2]
##  igraph                 1.2.5      2020-03-19 [1]
##  IRanges                2.20.2     2020-01-13 [2]
##  irlba                  2.3.3      2019-02-05 [2]
##  jsonlite               1.7.1      2020-09-07 [1]
##  KernSmooth             2.23-16    2019-10-15 [3]
##  knitr                  1.26       2019-11-12 [2]
##  labeling               0.3        2014-08-23 [2]
##  laeken                 0.5.1      2020-02-05 [1]
##  lattice                0.20-38    2018-11-04 [3]
##  lazyeval               0.2.2      2019-03-15 [2]
##  leiden                 0.3.1      2019-07-23 [2]
##  lifecycle              0.2.0      2020-03-06 [1]
##  listenv                0.8.0      2019-12-05 [2]
##  lmtest                 0.9-37     2019-04-30 [2]
##  lsei                   1.2-0      2017-10-23 [2]
##  lubridate              1.7.9.2    2020-11-13 [1]
##  magrittr             * 2.0.1      2020-11-17 [1]
##  MASS                   7.3-51.5   2019-12-20 [3]
##  Matrix                 1.2-18     2019-11-27 [3]
##  matrixStats            0.56.0     2020-03-13 [1]
##  memoise                1.1.0      2017-04-21 [2]
##  metap                  1.2        2019-12-08 [2]
##  mnormt                 1.5-5      2016-10-15 [2]
##  modelr                 0.1.8      2020-05-19 [1]
##  multcomp               1.4-12     2020-01-10 [2]
##  multtest               2.42.0     2019-10-29 [2]
##  munsell                0.5.0      2018-06-12 [2]
##  mutoss                 0.1-12     2017-12-04 [2]
##  mvtnorm                1.0-12     2020-01-09 [2]
##  nlme                   3.1-143    2019-12-10 [3]
##  nnet                   7.3-12     2016-02-02 [3]
##  npsurv                 0.4-0      2017-10-14 [2]
##  numDeriv               2016.8-1.1 2019-06-06 [2]
##  openxlsx               4.1.5      2020-05-06 [1]
##  pbapply                1.4-2      2019-08-31 [2]
##  pcaMethods             1.78.0     2019-10-29 [2]
##  pillar                 1.4.6      2020-07-10 [1]
##  pkgbuild               1.0.6      2019-10-09 [2]
##  pkgconfig              2.0.3      2019-09-22 [1]
##  pkgload                1.0.2      2018-10-29 [2]
##  plotly                 4.9.1      2019-11-07 [2]
##  plotrix                3.7-7      2019-12-05 [2]
##  plyr                   1.8.5      2019-12-10 [2]
##  png                    0.1-7      2013-12-03 [2]
##  prettyunits            1.1.1      2020-01-24 [1]
##  processx               3.4.2      2020-02-09 [1]
##  progeny              * 1.11.3     2020-10-22 [1]
##  proxy                  0.4-24     2020-04-25 [1]
##  ps                     1.3.2      2020-02-13 [1]
##  purrr                * 0.3.4      2020-04-17 [1]
##  R.methodsS3            1.7.1      2016-02-16 [2]
##  R.oo                   1.23.0     2019-11-03 [2]
##  R.utils                2.9.2      2019-12-08 [2]
##  R6                     2.4.1      2019-11-12 [1]
##  ranger                 0.12.1     2020-01-10 [1]
##  RANN                   2.6.1      2019-01-08 [2]
##  rappdirs               0.3.1      2016-03-28 [2]
##  RColorBrewer           1.1-2      2014-12-07 [2]
##  Rcpp                   1.0.4      2020-03-17 [1]
##  RcppAnnoy              0.0.16     2020-03-08 [1]
##  RcppEigen              0.3.3.7.0  2019-11-16 [2]
##  RcppHNSW               0.2.0      2019-09-20 [2]
##  RcppParallel           4.4.4      2019-09-27 [2]
##  RCurl                  1.98-1.1   2020-01-19 [1]
##  Rdpack                 0.11-1     2019-12-14 [2]
##  readr                * 1.4.0      2020-10-05 [1]
##  readxl               * 1.3.1      2019-03-13 [2]
##  rematch                1.0.1      2016-04-21 [2]
##  remotes                2.1.0      2019-06-24 [2]
##  reprex                 0.3.0      2019-05-16 [2]
##  reshape2               1.4.3      2017-12-11 [2]
##  reticulate             1.14       2019-12-17 [2]
##  rio                    0.5.16     2018-11-26 [1]
##  rlang                  0.4.8      2020-10-08 [1]
##  rmarkdown              2.0        2019-12-12 [2]
##  robustbase             0.93-6     2020-03-23 [1]
##  ROCR                   1.0-7      2015-03-26 [2]
##  rprojroot              1.3-2      2018-01-03 [2]
##  RSpectra               0.16-0     2019-12-01 [2]
##  rstudioapi             0.11       2020-02-07 [1]
##  rsvd                   1.0.3      2020-02-17 [1]
##  Rtsne                  0.15       2018-11-10 [2]
##  rvest                  0.3.6      2020-07-25 [1]
##  S4Vectors              0.24.2     2020-01-13 [2]
##  sandwich               2.5-1      2019-04-06 [2]
##  scales                 1.1.0      2019-11-18 [2]
##  scatterplot3d          0.3-41     2018-03-14 [1]
##  sctransform            0.2.1      2019-12-17 [2]
##  SDMTools               1.1-221.2  2019-11-30 [2]
##  sessioninfo            1.1.1      2018-11-05 [2]
##  Seurat               * 3.1.2      2019-12-12 [2]
##  SingleCellExperiment   1.8.0      2019-10-29 [2]
##  smoother               1.1        2015-04-16 [1]
##  sn                     1.5-4      2019-05-14 [2]
##  sp                     1.4-2      2020-05-20 [1]
##  stringi                1.5.3      2020-09-09 [1]
##  stringr              * 1.4.0      2019-02-10 [1]
##  SummarizedExperiment   1.16.1     2019-12-19 [2]
##  survival               3.1-8      2019-12-03 [3]
##  testthat               2.3.2      2020-03-02 [1]
##  TFisher                0.2.0      2018-03-21 [2]
##  TH.data                1.0-10     2019-01-21 [2]
##  tibble               * 3.0.4      2020-10-12 [1]
##  tidyr                * 1.1.2      2020-08-27 [1]
##  tidyselect             1.1.0      2020-05-11 [1]
##  tidyverse            * 1.3.0      2019-11-21 [2]
##  tsne                   0.1-3      2016-07-15 [2]
##  TTR                    0.23-6     2019-12-15 [1]
##  usethis                1.5.1      2019-07-04 [2]
##  uwot                   0.1.5      2019-12-04 [2]
##  vcd                    1.4-7      2020-04-02 [1]
##  vctrs                  0.3.5      2020-11-17 [1]
##  VIM                    6.0.0      2020-05-08 [1]
##  viridis              * 0.5.1      2018-03-29 [2]
##  viridisLite          * 0.3.0      2018-02-01 [2]
##  withr                  2.3.0      2020-09-22 [1]
##  xfun                   0.12       2020-01-13 [2]
##  xml2                   1.3.2      2020-04-23 [1]
##  xts                    0.12-0     2020-01-19 [1]
##  XVector                0.26.0     2019-10-29 [2]
##  yaml                   2.2.1      2020-02-01 [1]
##  zip                    2.0.4      2019-09-01 [1]
##  zlibbioc               1.32.0     2019-10-29 [2]
##  zoo                    1.8-7      2020-01-10 [2]
##  source                                 
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.3)                         
##  CRAN (R 3.6.2)                         
##  Bioconductor                           
##  Bioconductor                           
##  Bioconductor                           
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  Github (clauswilke/colorblindr@1ac3d4d)
##  R-Forge (R 3.6.2)                      
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.3)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  Bioconductor                           
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  Bioconductor                           
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.3)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.3)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  Bioconductor                           
##  Bioconductor                           
##  Bioconductor                           
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.3)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  Bioconductor                           
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  Bioconductor                           
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  Bioconductor                           
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.3)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  Github (saezlab/progeny@94bea1c)       
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.3)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.3)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  Bioconductor                           
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  Bioconductor                           
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.3)                         
##  CRAN (R 3.6.3)                         
##  Bioconductor                           
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.3)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  Bioconductor                           
##  CRAN (R 3.6.2)                         
##  CRAN (R 3.6.2)                         
##  Bioconductor                           
##  CRAN (R 3.6.2)                         
## 
## [1] /home/uhlitzf/R/lib
## [2] /usr/local/lib/R/site-library
## [3] /usr/local/lib/R/library

MSK SPECTRUM freeze major subset deep dives

Florian Uhlitz

2021-03-01

1 Cluster markers

1.1 Major Myeloid.super markers for cell assign

1.2 Subtype currated markers

1.3 Subtype cluster markers

2 Clusters

2.1 sizes

2.2 UMAP

3 Filtering out doublet and dissociated clusters

3.1 UMAP

3.2 add module scores and pathway scores

3.3 marker heatmap

3.4 composition

3.4.1 per site

3.4.2 per sample

3.4.3 site specific cluster enrichment

3.4.4 macrophage diffusion maps

3.4.5 UMAP & DC

4 session info